Recovery, Relapse, and Episodes of Default in The Management of Acute Malnutrition in Children in Humanitarian Emergencies: A Systematic Review

MARCH 2017
systematic review
Humanitarian Evidence Programme
RECOVERY, RELAPSE AND EPISODES

OF DEFAULT IN THE MANAGEMENT OF
ACUTE MALNUTRITION IN CHILDREN IN
HUMANITARIAN EMERGENCIES
About this systematic review
This is an independent systematic review commissioned by the Humanitarian Evidence
Programme, a partnership between Oxfam GB and Feinstein International Center at the
Friedman School of Nutrition Science and Policy, Tufts University. It was funded by aid from
the United Kingdom (UK) government through the Humanitarian Innovation and Evidence
(HIEP) Programme at the Department for International Development. The views and
opinions expressed herein are those of the authors and do not necessarily represent those
of Oxfam, Feinstein International Center or the UK government.
About the research team

This systematic review was led by Robert Akparibo and Andrew Lee at the University of
Sheffield. The review team included Andrew Booth, Janet Harris, Helen B. Woods, Lindsay
Blank, Michelle Holdsworth at the University of Sheffield, with Seth Adu-Afarwuah, Mark
Manary and Tanya Khara acting as scientific advisers to the team. An independent revision
was conducted by a team at the Friedman School of Nutrition Science and Policy, Tufts
University.
Corresponding author: Robert Akparibo, r.akparibo@sheffield.ac.uk.
The initial database and website searches took place between 1 November 2015 and
31 March 2016.
Citation
Akparibo, R., Lee, A.C.K. and Booth, A. (2017). Recovery, relapse and episodes of default in
the management of acute malnutrition in children in humanitarian emergencies: A systematic
review. Humanitarian Evidence Programme. Oxford: Oxfam GB.
Acknowledgments
We wish to thank Oxfam GB, Feinstein International Center and the UK Department for
International Development for funding this review. Our sincere thanks go to the revision team
and everyone who contributed thoughts, reviewed drafts and shared documents during the
development of this review. We particularly want to thank Tanya Khara for her technical advice
and Eleanor Ott, Roxanne Krystalli and Lisa Walmsley for providing editorial support. We
would also like to thank the organizations and individuals who responded to our requests for
their programme reports.
Series editors
This report forms part of a series of humanitarian evidence syntheses and systematic
reviews covering child protection, market support, mental health, nutrition, pastoralist
livelihoods, shelter, urban contexts and water, sanitation and hygiene. The reports and their
corresponding protocols can be found at:
https://www.gov.uk/dfid-research-outputs
http://fic.tufts.edu/research-item/the-humanitarian-evidence-program/
http://policy-practice.oxfam.org.uk/our-work/humanitarian/humanitarian-evidence-
programme.
The series editors are: Roxanne Krystalli, Eleanor Ott and Lisa Walmsley.
Photo credit
Azel, Niger, March 2015. Abbie Trayler-Smith/Oxfam.
Copyright Oxfam GB 2017

This publication is subject to copyright but the text may be used free of charge for the
purposes of advocacy, campaigning, education and research, provided that the source is
acknowledged in full. The copyright holder requests that all such use be registered with them
for impact assessment purposes. For copying in any other circumstances, or for re-use in
other publications, or for translation or adaptation, permission must be secured and a fee
may be charged. Email: lwalmsley1@ght.oxfam.org
CONTENTS
0 EXECUTIVE SUMMARY I
1 INTRODUCTION 1
1.1 Background 1
1.2 Rationale and purpose 2
1.3 Research questions 3
0.1 Population, intervention, control and outcome (PICO) parameters 3
2 METHODS 5
2.1 Protocol development 5
2.2 Implementation of the search strategy 5
2.3 Modifications from the published protocol 5
2.4 Document screening 6
2.5 Evidence appraisal 7
2.6 Data extraction and evidence synthesis 7
3 RESULTS 8
3.1 Search results 8
3.2 Characteristics of included studies 8
3.3 Assessment of study quality 9
3.4 Narrative synthesis of results 9
4 CONCLUSION 12
4.1 Limitations of the review 12
4.2 influence of Operational realities on programming and research 13
4.3 Age and sex data 13
4.4 Recommendations for future research 13
APPENDIX 1: DATABASE SEARCH STRATEGY STRINGS 15

Initial search strategy/outputs of scoping search 15
Final search strategy 15
Grey literature search 16
APPENDIX 2: ASSESSMENT OF INCLUDED STUDIES 17

Quality assessment of included quantitative case control and observational studies and
grey literature 17
Quality assessment of included quantitative RCTs and grey literature 18
Quality assessment of included qualitative studies and grey literature 18
APPENDIX 3: INCLUDED PUBLICATIONS REPORTED RESULTS 19

Characteristics of included mixed methods and observational cohort studies 19
Characteristics of included grey literature programme evaluation reports 21
Characteristics of included randomized controlled trials 22
REFERENCES 24
Included publications 24
Other publications 25
ABBREVIATIONS
CASP Critical Appraisal Skills Programme
CBTC Community-based therapeutic care
CMAM Community-based management of acute malnutrition
CRED Centre for Research on the Epidemiology of Disasters
CSB Corn/soy blend
CTC Community therapeutic care
DFID Department for International Development
FBF Fortified blended food
FS Fortified spread
GAM Global acute malnutrition
GNC Global Nutrition Cluster
HEP Humanitarian Evidence Programme
ICRC International Committee of the Red Cross
IMAM Integrated management of acute malnutrition
LNS Lipid-based nutrition supplement
MAM Moderate acute malnutrition
MSF Mdecins Sans Frontires
MUAC Mid-upper arm circumference
NCHS National Centre for Health Statistics
OTP Outpatient therapeutic feeding programme
RCT Randomized controlled trial
RUF Ready-to-use food
RUSF Ready-to-use supplementary food
RUTF Ready-to-use therapeutic food
SAM Severe acute malnutrition
SFP Supplementary feeding programme
TFC Therapeutic feeding centre
TSFP Targeted supplementary feeding programme
UNICEF United Nations Childrens Fund
WHO World Health Organization
WHZ Weight-for-height z-score
0 EXECUTIVE SUMMARY
This systematic review, commissioned by the Humanitarian Evidence Programme (HEP)
and carried out by a research team from the University of Sheffield, represents the first
attempt to apply systematic review methodology to establish the relationships between
recovery and relapse and between default rates and repeated episodes of default or relapse
in the management of acute malnutrition in children in humanitarian emergencies in low- and
1
middle-income countries.
Definitions and scope

Severe acute malnutrition (SAM, or severe wasting) and moderate acute malnutrition (MAM, or moderate
wasting) affect 52 million children under five years of age around the globe. This systematic review seeks to
establish whether there is a relationship between recovery and relapse or a relationship between default rates
and/or repeated episodes of default or relapse following treatment for SAM and MAM in children aged 659
months in humanitarian emergencies. The review also seeks to determine the reasons for default and relapse in
the same population.
For the purposes of this review:
Definitions of SAM and MAM are based on the National Centre for Health Statistics (NCHS) child growth
standard where studies are published before or during 2006, and the World Health Organization (WHO)
Child Growth Standards where published after 2006.
Humanitarian emergencies are defined as major incidents that threaten human life and public health (GNC,
2014; UNICEF, 2014), including protracted conflicts, flooding, earthquakes and other natural disasters. The
team used data from ReliefWeb and the Centre for Research on the Epidemiology of Disasters (CRED) to
identify the existence of a humanitarian context, which might include seasonal spikes in malnutrition, a
declared food crisis, natural disasters or disease outbreaks that affected nutritional status.
Recovery rate means the proportion of children who are cured through treatment in acute malnutrition
programmes.
Default rate means the proportion of children absent from treatment for two consecutive sessions.
Absence and/or repeated absence means the proportion of children who were absent from treatment/the
number of absences recorded.
Return default rate/repeated default episodes means the proportion of children who re-enrolled into
treatment after defaulting/the number of times they re-enrolled.
Relapse rate/repeated relapse episodes means the proportion of children who re-enrolled after they had
recovered and been discharged.
Time to recover means length of time between admission and discharge.
The research team:

mapped and documented existing research;
synthesized the evidence in response to two key research questions:
the relationship between recovery and relapse; and the relationship between relapse
and default or return default/episodes of default
the reasons for default and relapse or return defaults/episodes of default; and
identified gaps in existing research and knowledge.
What evidence was eligible for review?

The research team identified a total of 9,574 articles, studies and programme reports relating
to acute malnutrition in its initial search of databases and websites. Following the removal of
duplicates, screening and quality appraisal, 24 articles were eligible for review. Of the 24
eligible studies, 23 focused on sub-Saharan Africa: eight were conducted in Malawi, five in
Ethiopia, three in Niger, three in Sudan and the remaining four in Angola, Chad, Kenya and
Sierra Leone. One study focused on Afghanistan. Most studies and programme reports
reported on quantitative outcomes and two contained both quantitative and qualitative
outcomes. The 22 quantitative studies included eight clinical efficacy and effectiveness trials
using randomized controlled designs, seven observational cohort studies and seven
programme evaluation reports.
1
HEP is a partnership between Oxfam GB and Feinstein International Center at the Friedman School of Nutrition Science and Policy,
Tufts University. It is funded by the United Kingdom governments Department for International Development (DFID) through the
Humanitarian Innovation and Evidence Programme (HIEP).
Recovery, relapse and episodes of default in the management of acute malnutrition in children in humanitarian emergencies ii
What are the findings?

What is the relationship between recovery and relapse, and between relapse and
default or return default/episodes of default?
Six of the 24 studies included in this review addressed the issue of relapse and/or reported
relapse rates (Bahwere et al., 2014; Ciliberto et al., 2005; Linneman et al., 2007; Querubin,
2006; Taylor, 2002; UNICEF, 2012). Figure 0.1 presents the relapse and default rates
reported for each of these studies. None of the studies addressed the relationship between
relapse and default or return default. This may be partly attributable to the relatively short
duration of most interventions and the fact that most of the studies did not include post-
intervention follow-up.
Figure 0.1: Relapse rates and relationship between relapse and default
Study Country Type Quality Findings

Bahwere et Malawi Randomized High Briefly mentions relapse but does not report rate.
al. (2014) controlled clinical
effectiveness trial
Ciliberto et al. Malawi Clinical Medium Higher default rate (9.8 percent) and lower
(2005) effectiveness trial relapse/mortality rate (8.7 percent) reported for
children receiving home-based care vs. those
receiving standard therapy (8.1 percent and 16.7
percent respectively).
Linneman et Malawi Observational High Average default rates of 7 percent for children
al. (2007) cohort study with SAM and 8 percent for children with MAM
reported. Relapse rates not reported separately
from non-recovery rates (3 percent and 4 percent
for children with SAM and MAM respectively).
Querubin Sudan Programme Low Default rate of 7 percent and no cases of relapse
(2006) evaluation report reported for children in home treatment group.
Taylor (2002) Sudan Programme Low Average default rate of 10.1 percent reported.
evaluation report Readmission rate approximately 1 percent of
total admissions.
UNICEF Kenya Programme Low Default rate of 12.9 percent and relapse rate of
(2012) evaluation report 3.2 percent reported for outpatient SAM
treatment. Default rate of 1.4 percent and relapse
rate of 6.1 percent reported for inpatient SAM
treatment. Default rate of 14.4 percent and
relapse rate of 3.7 percent reported for MAM
treatment.
What are the reasons for default and relapse or return defaults/episodes of default?
Default data was reported in each of the 24 included studies, but reasons for default were
not always cited (see Figure 0.2). Relapse was reported in six of the included studies but
only one study discussed possible reasons for relapse. A UNICEF programme evaluation
report of an integrated management of acute malnutrition (IMAM) programme in Kenya
(2012; low quality) suggests that the lack of a follow-up system to track children treated for
SAM or MAM, together with a lack of encouragement to return for outpatient treatment, may
have contributed to relapses, as well as the sharing of ready-to-use therapeutic food (RUTF)
among siblings and other non-admitted children.
Recovery, relapse and episodes of default in the management of acute malnutrition in children in humanitarian emergencies iii
Figure 0.2: Reasons for default reported in studies
Study Country Type Quality
Family illness
priorities
Other household
Travel distance
sensitization
Poor community
Poor follow-up
Other issues
offered or not applicable

No reason for default
Amthor et al. (2009) Malawi Observational High X
cohort study
Bahwere et al. (2014) Malawi Randomized High X
controlled trial
Belachew and Nekatibeb Ethiopia Mixed methods Medium X X X
(2007) study
Chaiken et al. (2006) Ethiopia Observational Low X
cohort study
Ciliberto et al. (2005) Malawi Randomized Medium X
controlled trial
Collins and Sadler (2002) Ethiopia Observational Medium X
cohort study
Deconinck (2004) Ethiopia Programme Low X
evaluation
report
Flax et al. (2009) Malawi Mixed methods Medium X
study
Gaboulaud et al. (2007) Niger Observational Low X
cohort study
Grellety et al. (2012) Niger Observational High X
cohort study
Huybregts et al. (2012) Chad Randomized Medium X
controlled trial
Isanaka et al. (2009) Niger Randomized Medium X
controlled trial
Karakochuk et al. (2012) Ethiopia Randomized High X
controlled trial
Lagrone et al. (2010) Malawi Observational Medium X
cohort study
Linneman et al. (2007) Malawi Observational High X*
cohort study
Lurqin (2003) Afghanistan Programme Low X X X
evaluation report
Matilsky et al. (2009) Malawi Randomized High X
controlled trial
Maust et al. (2015) Sierra Leone Randomized Medium X
controlled trial
Morgan et al. (2015) Angola Programme Low X
evaluation report
Oakley et al. (2010) Malawi Randomized High X**
controlled trial
Querubin (2006) Sudan Programme Low X
evaluation report
Taylor (2002) Sudan Programme Low X
evaluation report
UNICEF (2012) Kenya Programme Low X
evaluation report
Walker (2004) Sudan Programme Low X
evaluation report
* Authors indicated that high default rates at two study locations could be attributable to undiagnosed HIV infection.
** Authors did not find sufficient evidence to determine the contribution to default rates of taste/colour differences between products.
Recovery, relapse and episodes of default in the management of acute malnutrition in children in humanitarian emergencies iv
What is the state of the evidence on acute malnutrition?

This review found little evidence on the impact of programmes implemented to manage
MAM and SAM in emergencies, which can be partly explained by the nature of interventions
conducted in humanitarian settings. Also, it found few studies that explored the issue of
relapse.
There is considerable heterogeneity in the evidence base owing to the diversity of study
types, types of intervention and settings in which the programmes were delivered. The
quality of the published literature, however, varies markedly:
Generally, the grey literature reports and qualitative studies included in the review lacked
crucial detail, such as details of their recruitment strategies.
The review of articles did identify the presence of significant barriers relating to the
acceptability of the intervention and implementation issues.
The detail presented in the mixed methods studies offers a more nuanced understanding
of the precise context and, indeed, suggests that some factors are generic and others,
even though present, may not have the same relative importance in a different context.
The lack of qualitative data on acute malnutrition in humanitarian contexts is striking,

with only two mixed methods studies eligible for inclusion in this review (Belachew and
Nekatibeb, 2007; Flax et al., 2009). However, it is possible that qualitative studies related to
treatment of acute malnutrition are better represented in nutritional emergencies, and were
therefore left out of this review. It is possible that nutritional emergencies exist in non-
humanitarian settings, often as a result of inequities related to poverty. For example, India
has the highest number of acutely malnourished children under five in the world
25,461,000 but this is not considered a humanitarian emergency.
Child-level data, disaggregated by age and gender, is routinely collected in the

management of acute malnutrition programmes, although this was not reported in the
included studies. This is a gap in understanding how management of acute malnutrition may
result in default and/or relapse differently in boys compared with girls. Given that boys are
more likely to be wasted than girls in humanitarian emergency settings (Wamani et al.,
2007), there could be a different type of programming modality for boys versus girls at
certain ages. Also, the lack of age breakdown is a gap in this review, since children aged 6
23 months are often the most wasted age group (UNICEF, 2013). Age category analysis can
inform how programmes can adapt to age-specific diet and psychosocial needs in order to
mitigate relapse and default, and improve recovery. Evidence on the relationship between
recovery and relapse following management of acute malnutrition in humanitarian
emergencies is sparse.
Overall conclusions
This review identified 24 publications and programme reports that reported recovery, relapse
and default and reasons for recovery and default in the management of acute malnutrition in
humanitarian emergencies. Assessment of these publications did not provide sufficient
information for conclusions to be drawn in relation to the review questions.
The relationship between recovery and relapse; and between relapse and default or
return default/episodes of default: The evidence related to this theme is inconclusive.
Reasons for default and relapse or return defaults/episodes of default: The strength of
evidence relating to this theme is limited.
This review provides further confirmation that RUTF used in an outpatient setting is effective
at promoting recovery from SAM and reducing mortality. It could not be established whether
default rates reported were lower according to the WHO 2013 protocol, which formally
included the community-based treatment of acute malnutrition using RUTF, compared with
other methods to treat acute malnutrition. Data relating to relapse is limited.
Recovery, relapse and episodes of default in the management of acute malnutrition in children in humanitarian emergencies v
Similarly, as reported in the studies, the use of ready-to-use supplementary foods (RUSFs)
to treat MAM, compared with other supplementary foods such as fortified blended foods
(FBFs) or corn/soy blends (CSBs), has the potential to improve nutritional recovery in
children suffering from MAM. However, the weight gain reported for children given RUSFs or
CSBs was small. No significant differences in mortality rates were reported for children who
received RUSF compared with children in CSB groups. It could not be concluded whether
RUSF could potentially improve relapse and default rates or reduce mortality as there was
insufficient robust data available to undertake this analysis.
Little evidence was found on the long-term impact of programmes implemented to manage
MAM and SAM in emergencies. Due to the nature of humanitarian emergencies, it is difficult
to conduct longitudinal data collection and trace the same children once they have been
discharged. The mortality rates reported in these studies ranged from 0.007 percent
(Isanaka, 2009) to 18.9 percent (Gaboulaud, 2006). Low mortality and relapse rates were
associated with access to skilled medical care, availability of essential drugs, greater
parental awareness of the consequences of illness and the need to seek available health
services early.
While reasons for default were sometimes cited including distance to travel to sites, family
illness and other commitments, other household priorities, poor community sensitization and
household follow-up none of the studies included in the review focused on defaulted
children as an objective.
Barriers to successful completion of acute malnutrition treatment were multi-factorial and

included geographic, cultural and socioeconomic obstacles to care (Guerrero et al., 2013).
For successful implementation of the WHO protocols for community-based management of
acute malnutrition (CMAM), emergency nutrition workers need to be able to adapt protocols
to each context.
Programmatic issues and operational challenges also vary between settings and are
powerful determinants of the efficacy of programmes. Additional factors for efficacious
programme implementation included the degree of integration into existing health systems,
the extent of service coverage, staff training and alignment of inpatient treatment with WHO
guidelines. However, there were pervasive problems such as resource constraints, the lack
of adequate supervision and monitoring, lack of adequate outreach to identify the most
vulnerable, dependence on external technical support, weak pipelines for specialized
nutritious foods and logistical constraints.
A number of research gaps were identified during this review. In particular, there is a need
for more evidence on default and relapse post-treatment; for in-depth analysis of contexts
where default and/or relapse rates have been historically high; for observational studies on
care practices for malnourished children during and after treatment; and for studies to
examine the long-term effects of SAM treatment, in particular the relationship between
wasting and stunting.
1 INTRODUCTION
1.1 BACKGROUND
Acute malnutrition is a major contributor to deaths in children under five years of age
worldwide (WHO and UNICEF, 2009; UNICEF, 2014). According to current estimates,
approximately 17 million children under the age of five suffer from severe acute malnutrition
(SAM) and 33 million from moderate acute malnutrition (MAM) globally (UNICEF, WHO and
World Bank, 2016). Nearly two-thirds of the burden of acute malnutrition is borne by South-
East Asia and sub-Saharan Africa (UNICEF, WHO and World Bank, 2012). South Sudan
has the highest level of global acute malnutrition (GAM) (23 percent) and SAM (10 percent),
with a GAM level of 15 percent regarded as the threshold for defining an emergency
(UNHCR, WFP, 2011).
Defining acute malnutrition in children aged 659 months

The classification of acute malnutrition has evolved, with improved evidence to define optimal growth. The
World Health Organization (WHO) Child Growth Standards were published in 2006 and were based on a global
sample of optimally fed infants from birth through five years of age. Before 2006, interventions implemented to
manage acute malnutrition classified children as acutely malnourished based on the US-based National Centre
for Health Statistics (NCHS) child growth standard or on the presence of bilateral pitting oedema (WHO, 1999).
These reference standards compared children under five with a reference population, many of whom were
artificially fed. The 2006 WHO Child Growth Standards are now used to identify a child as being malnourished
(WHO, 2013). The table below outlines the differences between the two systems of classification. Mid-upper
arm circumference (MUAC) was added as an independent admission criterion in the 2013 update.
Classification Severe acute malnutrition (SAM) Moderate acute malnutrition (MAM)

NCHS <70% median <80% median
<-3 z-score >=-3 and <-2 z-score
Oedema Oedema
WHO, 2013 <-3 z-score or oedema >=-3 and <-2 z-score or oedema
MUAC: <115mm MUAC: <125mm
MUAC is more predictive of mortality and tends to select more children who are younger, stunted and female
(Isanaka et al., 2015). MUAC has been effectively used in most community-based management of acute
malnutrition (CMAM) programmes to screen and refer children for appropriate treatment (Myatt et al., 2006).
MUAC is also increasingly used as an independent admission criterion for treatment and discharge. The studies
included in this review may use different indicators to define acute malnutrition, depending on when they were
conducted. This may have implications for study outcomes, but whenever possible comparisons have been
made between equivalent indicators.
In the aftermath of a humanitarian emergency, a breakdown in health services, destruction

of food sources and markets and loss of infrastructure can result in an increase in acute
malnutrition (Bagchi et al., 2004). Humanitarian emergencies may also disrupt cultivation
and harvest, negatively impacting the food security status of an affected population. The risk
of disease outbreaks increases with displacement, overcrowding and a breakdown in
sanitation and hygiene, which also increases the risk of undernutrition (Brundtland, 2000).
Compromised infant and young child feeding practices, resulting from displacement,
stress/trauma, lack of complementary foods and a breakdown in social structure, are also
common during a humanitarian emergency. Nutritional emergencies can also exist in non-
humanitarian settings, often as a result of inequities related to poverty. For example, India
has the highest number of acutely malnourished children aged under five in the world
25,461,000 but this is not considered a humanitarian emergency. Most of the worlds
acutely malnourished children (33.9 million) live in Asia (UNICEF, WHO and World Bank,
2015). The UN has highlighted that the wasting rate in Southern Asia is approaching a
critical public health emergency, at 14.1 percent GAM in its most recent estimate (UNICEF,
WHO and World Bank, 2016).
The WHO Guideline, Management of severe malnutrition: a manual for physicians and other
senior health workers, was first published in 1999 (WHO, 1999) and has guided the inpatient
management of acute malnutrition in hospital or specialized nutrition rehabilitation wards since
then. In 2013, WHO published updates on the management of SAM in infants and children
Recovery, relapse and episodes of default in the management of acute malnutrition in children in humanitarian emergencies 2
(WHO, 2013), which include CMAM interventions for children without medical complications.
There are no global guidelines for the management of MAM, but there are country-level
national guidelines for acute malnutrition which incorporate the treatment of MAM and a Global
Nutrition Cluster (GNC) guidance document from the MAM taskforce (Force, 2012). CMAM
programmes exist in over 70 countries worldwide (UNICEF, 2014) and are the standard
intervention for responding to acute malnutrition during a humanitarian crisis.
Management of acute malnutrition

The WHO guideline, Management of severe malnutrition: a manual for physicians and other senior health
workers, published in 1999, outlines the phases of inpatient care for children: initial stabilization (Phase I),
nutritional rehabilitation (Phase II) and follow-up. Phase I requires the use of F-75 therapeutic milk and the
treatment of life-threatening conditions such as shock and dehydration. Phase II involves continued nutritional
rehabilitation with F-100 therapeutic milk until the discharge weight is achieved. After discharge, health workers
are supposed to conduct regular follow-up visits to the homes of treated children to support and counsel their
care-givers, with the aim of preventing relapse (WHO, 1999).
The CMAM model was introduced to address both SAM and MAM at the community and household levels.
Different terminologies have been used to describe the CMAM model, including community therapeutic care
(CTC), integrated management of acute malnutrition (IMAM), community-based therapeutic care (CBTC) and
ambulatory care for the treatment of acute malnutrition. This review uses CMAM to refer to the community
management of acute malnutrition. CMAM programmes use essential medicines and RUTFs to treat SAM in
outpatient therapeutic feeding centres and maintain stabilization centres for children who have SAM with
medical complications and require 24-hour care.
Another component of CMAM is the treatment of MAM. While MAM treatment is not included in the WHO 2013
protocol, there are different approaches in national guidelines. MAM treatment through targeted supplementary
feeding programmes (TSFPs) employs fortified blended foods (FBFs), such as corn-soy blends (CSBs) or
SuperCereal Plus, which are fortified with micronutrient pre-mixes and may include vegetable oil, whey or
skimmed milk powder (Webb et al., 2011). Ready-to-use supplementary foods (RUSFs) have now been
designed as an alternative to FBFs to treat MAM (UNICEF, 2014). These are energy-dense fortified foods that
are packaged in a convenient way and are ready for use, i.e. they do not require cooking or preparation at home
before use, unlike FBFs (Nutriset, 2008).
The CMAM model was developed in the late 1990s and early 2000s, and early CMAM programmes achieved
nutritional recovery rates of almost 80 percent, with low case fatality rates of <5 percent (Collins et al., 2006).
Other clinical trials (Diop et al., 2004; Manary et al., 2004; Ndekha et al., 2005) as well as descriptive studies
(Webb et al., 2014) conducted in emergency nutrition situations have all demonstrated that the CMAM approach
can achieve successful outcomes in nutritional recovery and promote child survival.
Performance of CMAM is measured through four main indicators: recovery the proportion of children aged 6
59 months who successfully recover; died the proportion of children who die during treatment; default the
proportion of children who default from (do not complete) treatment; relapse the proportion of children who
complete treatment, but develop SAM and are readmitted for further treatment. The recovery rates reported in
most studies have been within the Sphere Projects Minimum Standards or indicators for therapeutic
effectiveness (Sphere, 2011).
Sphere Project: minimum standards for measuring programme effectiveness
Indicators Proportions
Proportion of children who die in the programme <10%
Proportion of children who recover in the programme >75%
Proportion of children who default from the programme <15%
Minimum rate of weight gain (g/kg/person/day) >8
Source: Sphere Project (2011)
1.2 RATIONALE AND PURPOSE

Previous systematic reviews have investigated the effectiveness of interventions to manage
acute malnutrition in children aged 659 months. Some have focused on comparing the
inpatient model with the community model (Lenters et al., 2013; Schooness et al., 2013;
Ashworth, 2006); some have assessed the appropriateness and efficacy of routine first-line
antibiotics provided during outpatient care for children with uncomplicated SAM (Alcoba et
al., 2013); and others have examined the effect of different ready-to-use foods (RUFs),
including cost-effectiveness (Bhutta et al., 2008; Gera, 2010). Picot and colleagues (2012)
reviewed various intervention studies that have been designed to treat infants and children
aged under five years with SAM. The most recent review (Suri et al., 2016) focusing on the
management of MAM compared RUTFs and FBFs to determine whether the inclusion of
dairy products modifies their comparative effectiveness.
Emergency intervention studies focusing on SAM treatment (Kerac et al., 2009; Giugliani et al.,
2010; Aguayo, 2013; Puett et al., 2013; Puett and Guerrero, 2015; Dale et al., 2013; Burza et
al., 2015), as well as the management of SAM and MAM (UNICEF, 2012; Somass et al.,
2015), have reported high rates of default and relapse. Evidence is sparse on the extent to
which children who recover during treatment for acute malnutrition in humanitarian
emergencies subsequently relapse. Further, little is known about how operational realities,
such as insecurity and access to a population, affect both the implementation of programmes
to treat acute malnutrition and research conducted on programmatic performance.
1.3 RESEARCH QUESTIONS

This systematic review was undertaken to examine outcomes beyond recovery. Specifically,
the review team sought to better understand how relapse and default are affected during
CMAM programmes in order to determine how programmes can be modified to improve
overall quality in humanitarian emergencies. A further aim was to synthesize findings from
qualitative research studies to explore the contexts in which acute malnutrition management
programmes have been implemented, in order to identify and describe how the
implementation context influences success or relapse and/or default.
The review investigated:

the relationship between recovery and relapse; and between relapse and default or return
default/episodes of default in children aged 659 months affected by humanitarian
emergencies
reasons for default and relapse or return defaults/episodes of default in children aged 6
59 months affected by humanitarian emergencies.
The findings of this review are expected to contribute to the growing literature on
interventions used to treat acute malnutrition in emergency settings, with the ultimate aim of
better informing programme and policy decisions.
0.1 POPULATION, INTERVENTION, CONTROL AND

OUTCOME (PICO) PARAMETERS
Population
The review included studies if they targeted children aged 659 months of age with
treatment of acute malnutrition in humanitarian emergencies. Studies that focused on the
management of acute malnutrition in infants less than six months of age and pregnant
women or lactating mothers were not included. Because of the different definitions and
classification criteria that have been used to identify children with acute malnutrition (see
Section 1.2), only studies that used standard and accepted definitions (Picot et al., 2012) of
SAM and MAM were included: i.e. definitions for MAM and SAM used by the authors should
be based on the NCHS child growth standard if studies were published before or during
2006 and on the WHO Child Growth Standards if studies were published after 2006.
Interventions/issues
The interventions considered for this review are those for the treatment of SAM or MAM
delivered in humanitarian emergencies. Humanitarian emergencies are defined as major
incidents that threaten human life and public health (WHO, 2010). These include protracted
conflicts, flooding, earthquakes and other natural disasters. Interventions provided to
manage/treat SAM in humanitarian emergencies should include inpatient care using
therapeutic milks such as F-75 or F-100 and RUTFs, and outpatient care using RUTFs.
Interventions provided to manage/treat MAM in humanitarian emergencies include those that
used FBFs e.g. SuperCereal, SuperCereal Plus, CSBs and/or oil-based RUSFs. Studies that
were based on blanket supplementary feeding provided to children to prevent MAM from
progressing to SAM were not considered. However, where studies state that the aim of the
supplementary feeding programme was to manage rather than prevent MAM they were
included in the review.
Comparison
This review compared relapse rates and default rates in children aged 659 months in the
management of acute malnutrition programmes in humanitarian emergencies.
Outcomes
Recovery rate: the proportion of children who are cured through treatment of acute
malnutrition programmes as defined by the authors based on standard criteria.
Default rate: the proportion of children who were absent from treatment for two
consecutive sessions.
Absence and/or repeated absence: the proportion of children who were absent from
treatment/the number of absences recorded.
Return default rate/repeated default episodes: the proportion of children who re-
enrolled into treatment after defaulting/number of times they re-enrolled.
Relapse rate/repeated relapse episodes: the proportion of children who re-enrolled

after they had recovered and been discharged.
Time to recover: the length of time between admission and discharge.

2 METHODS
2.1 PROTOCOL DEVELOPMENT
A protocol was developed and used to guide the review (Akparibo et al., 2016), setting out
the problem and the rationale for the review. It also provides a complete description of the
proposed methods for the review. The protocol was peer-reviewed and revised before the
review began. The full review protocol is available online at: http://fic.tufts.edu/assets/OX-
HEP-Malnutrition-Print.pdf
2.2 IMPLEMENTATION OF THE SEARCH STRATEGY

This systematic review used a mixed methods systematic literature review approach to
identify and include various types of literature: quantitative, qualitative and mixed methods
studies, as well as grey literature such as programme reports. The search was conducted in
the following databases: Medline and Medline In-Process; Embase; CINAHL; Science
Citation Index Expanded, Conference Proceeding Citation Index Science, and Book
Citation Index Science (all via Web of Science); regional literature databases: African
Index Medicus, Index Medicus for the Eastern Mediterranean Region (IMEMR), Latin
American and Caribbean Health Sciences Literature (LILACS), the Western Pacific Regional
Index Medicus (WPRIM) and Index Medicus for the South-East Asian Region (IMSEAR); the
Cochrane Central Register of Controlled Trials (CENTRAL) and BIOSIS. In addition, the
research team used cluster-searching techniques (Booth et al., 2013) to enable the retrieval
of sibling studies (those closely associated with previously identified randomized controlled
trials (RCTs) and observational studies) to broaden the collective understanding of the
context in which these studies were conducted.
Grey literature, including programme reports, was identified from websites of relevant
governments and non-governmental organizations (NGOs), via Google Scholar and by
engaging with stakeholders in the nutrition field. Grey literature was also collected by
contacting humanitarian organizations or individuals to ask if they had unpublished reports
relevant to the review.
Studies published prior to 1980 were excluded, as were any not published in English. The
rationale for the date limitation was that the standardized diagnostic criteria for classifying
SAM and MAM for treatment, as well as the development of tailored interventions to address
these conditions, occurred in the early 1980s. Very few of these studies fall within the
framework of humanitarian emergencies caused by protracted conflicts, flooding and
earthquakes. We emailed the authors of studies that were unclear to ask them to clarify
whether these were humanitarian emergency or non-humanitarian emergency studies.
2.3 MODIFICATIONS FROM THE PUBLISHED PROTOCOL

The first modification from the original protocol was the decision made, in consultation with
the review commissioners, not to address one of its specific objectives through the
systematic review process the effect of treatment on mortality risk among children
suffering from SAM and/or MAM (Akparibo et al., 2016). It was suggested in peer reviews
and by the systematic review expert board that this question had already been sufficiently
addressed. Previous reviews (Lenters et al., 2013; Ashworth et al., 2006) have highlighted
that the 1999 inpatient guidelines have resulted in improved nutritional recovery rates
(compared with non-standard care) and have led to significant reductions in mortality rates
(Briend et al., 1999; Collins, 2004).
Revisions following peer review

The second modification was a re-screening of included articles for a stricter application of
including only studies that were conducted in humanitarian emergencies. The protocol
defined humanitarian emergencies as:
major incidents that threaten human life and public health. These include protracted
conflicts, flooding, earthquakes and other natural disasters, and nutrition emergencies
when prevalence rates of malnutrition have reached a crisis level (>=15%), according
to the WHO crisis classification using global acute malnutrition. The latter in particular
may include government-led interventions designed to manage acute malnutrition in
such a crisis situation. (Akparibo et al., 2016)
Peer review feedback indicated that not all selected studies were from humanitarian
emergencies. To address this feedback, studies were also included if they did not take place
in a humanitarian setting but their objectives explicitly stated that the research was intended
to improve management of acute malnutrition programmes in humanitarian emergencies.
Following peer review, two sources (ReliefWeb and the Centre for Research on the
Epidemiology of Disasters (CRED)) were used to verify that a humanitarian emergency was
occurring at the time data was collected for each study. These databases were also used to
verify if the study had an objective to improve humanitarian response in the study area.
Considerations for the search in the databases included seasonal spikes in malnutrition, a
declared food crisis and natural disasters or disease outbreaks that affected nutrition status.
Articles that did not meet these criteria were not included in the review. It was determined
that 24 out of 69 studies referred to strictly humanitarian emergency settings.
2.4 DOCUMENT SCREENING

Guided by the inclusion or eligibility criteria, as well as purpose-specific checklists, all the
studies identified during the search were screened in two stages.
Stage 1
One reviewer screened all the titles of the studies initially identified for potentially relevant
studies, and a subject specialist checked this.
Stage 2
Four reviewers then conducted a further screening of the abstracts and executive
summaries of articles selected after initial screening in Stage 1 (n=1,521), based on the
article titles. The four reviewers were paired, and each pair was assigned a test set of 25
papers and reports to review in order to establish a consensual selection process.
The relevance of each paper was coded and cross-checked by a subject specialist. Further,
the quality of each study was thoroughly assessed and graded using standard checklists. To
minimize inter-observer variability in the screening and coding process, several measures
were instituted. For instance, before any coding was started, the subject specialist gave the
review team an orientation tutorial on the management of MAM and SAM in humanitarian
emergencies and the type of information required to address the review question. The
review team drew upon training materials developed by UNICEF, as well as a PowerPoint
presentation received from a nutrition specialist as resource or reference materials. We held
team meetings to resolve uncertainties or disagreements between reviewers, where
necessary.
2.5 EVIDENCE APPRAISAL

From the outset, we anticipated that studies to be included in the review would be based on
different methodological designs. The review team followed the Cochrane guidance for the
critical appraisal of qualitative research and selected the Critical Appraisal Skills Programme
(CASP) checklist for qualitative studies (CASP, 2013). CASP checklists (www.casp-uk.net)
were used to assess the quality of quantitative and systematic review studies as there are
separate checklists for these study types. Two reviewers applied CASP checklists to assess
and grade the quality of the included studies. They screened the full text of the filtered sets
of quantitative articles, and another two reviewers cross-checked their selections. The
studies were graded as high, moderate, low or very low quality. In accordance with recent
empirical research, we decided that the quality assessment process would not lead to the
exclusion of papers. Instead, it would be used to moderate the strength of evidence for the
thematic findings (Carroll et al., 2011). During revision, ReliefWeb and CRED databases
were used to classify articles from the search by context, either humanitarian or
development. Those articles that did not meet the revised, stricter interpretation of the review
criteria were omitted from the final review.
Figure 2.1: CASP checklist questions
CASP checklist for RCT quality CASP checklist for cohort study CASP checklist quality
assessment questions quality assessment questions assessment questions for
qualitative studies
Clearly focused issue Clearly focused issue Clear statement of aims
Randomized Recruitment acceptable Qualitative methodology
Patients/study personnel/health Exposure accurately measured appropriate
workers blinded Outcomes accurately measured Appropriate research design
Groups similar at start of trial Confounders identified Appropriate recruitment
Treated equally strategy
Follow-up long enough
Results precise Appropriate data collection
Rates or proportions reported
Can results be applied? Relationship between
Results precise researcher and participants
Clinically important outcomes Results believable Ethical issues considered
Benefits worth harms/cost Can results be applied? Data analysis rigorous
Fits with other evidence Findings clear
Provides implications for Value of research
practice
2.6 DATA EXTRACTION AND EVIDENCE SYNTHESIS

Methods of quantitative data synthesis were largely informed by the nature and type of
studies identified and included in the review. We did not conduct a meta-analysis of included
homogenous studies, since previous reviews (Gera, 2010; Schooness et al., 2013; Lenters
et al., 2013) have already done this. Instead, a narrative synthesis of the data was
performed, with a tabulation of the results.
3 RESULTS
3.1 SEARCH RESULTS
The PRISMA diagram presented in Figure 3.1 illustrates the screening process. A total of
9,574 studies and programme reports were retrieved. After duplicates were removed, 7,693
citations were left to be screened. An initial title screening resulted in 6,172 citations being
excluded. An abstract screening resulted in the exclusion of 1,419 additional studies. A full
text review was conducted of the remaining 102 studies, which resulted in the exclusion of
33 studies. To address peer review feedback, 45 studies that did not meet the humanitarian
crisis criteria were excluded. A total of 24 studies were therefore included in the review.
Figure 3.1: PRISMA diagram of included studies
Academic databases Grey literature Personal contact Other

research search with stakeholders searches
9,492 Citation(s) 56 Citation(s) 7 Citation(s) 7 Citation(s)
7,693
Non-duplicate
Citations screened
Inclusion/ 7,591 Articles

exclusion criteria excluded after
applied title/abstract screen
102 Articles
retrieved
Inclusion/ 33 Articles 45 Articles

exclusion criteria excluded after excluded by
applied full text screen review team
24 articles
included
3.2 CHARACTERISTICS OF INCLUDED STUDIES

A total of 24 articles were included in the review, based on the criteria set out in Section 2.
One study was conducted in Afghanistan and 23 were conducted in sub-Saharan Africa:
eight in Malawi, five in Ethiopia, three in Niger, three in Sudan and one each in Angola,
Chad, Kenya and Sierra Leone.
Of the 24 studies/programme reports eligible for inclusion, 22 reported on quantitative

outcomes and two on both quantitative and qualitative outcomes (mixed methods). The
quantitative studies were clinical efficacy and effectiveness trials using randomized
controlled designs (n=8), observational cohort studies found in journal articles (n=7) and
programme evaluation reports found in grey literature (n=7). The sample sizes ranged from
170 to 5,799 children for studies focusing on SAM, 1,038 to 23,865 children for studies
focusing on MAM and 7,028 to 97,692 children for studies focusing on both SAM and MAM.
The mixed methods studies examined issues relating to the implementation and
acceptability of programmes to combat acute malnutrition. Data collection methods
employed in the mixed methods studies included interviews (Belachew and Nekatibeb, 2007;
Flax et al., 2009) and documentary analysis, focus groups and observations (Belachew and
Nekatibeb, 2007). We identified five reviews, but only extracted data from their included
papers if they were relevant to our review.
3.3 ASSESSMENT OF STUDY QUALITY

The studies were evaluated based on various CASP checklists (CASP, 2013). Overall, the
included studies were categorized as high, medium, or low quality based on these checklists
(see Appendix 2). The observational cohort studies (n=7) and programme evaluation reports
(n=7) were evaluated based on the CASP Cohort Study Checklist. Three of the
observational studies were assessed to be of high quality (Amthor et al., 2009; Grellety et
al., 2012; Linneman et al., 2007); two were assessed to be of medium quality (Collins and
Sadler, 2002; Lagrone et al., 2010); and two were assessed to be of low quality (Chaiken et
al., 2006; Gaboulaud et al., 2006). All seven of the programme evaluation reports were
assessed to be of low quality (Deconinck, 2004; Lurqin, 2003; Morgan et al., 2015;
Querubin, 2006; Taylor, 2002; UNICEF, 2012; Walker, 2004). The quality of the RCTs (n=8)
was evaluated based on the CASP RCT Checklist. Four of these studies were assessed to
be of high quality (Bahwere et al., 2014; Karakochuk et al., 2012; Matilsky et al., 2009;
Oakley et al., 2010) and four were assessed to be of medium quality (Ciliberto et al., 2005;
Huybregts et al., 2012; Isanaka et al., 2009; Maust et al., 2015). The qualitative component
of the mixed methods studies (n=2) was evaluated based on the CASP Qualitative Checklist
and was assessed to be of medium quality (Belachew and Nekatibeb, 2007; Flax et al.,
2009).
Particular concerns related to the adequacy of descriptions of the recruitment strategy used
in the included studies. The extent to which participants expressed the range of opinions
held by the population in general was unclear. Little detail was provided of ethical
considerations with, at best, a cursory mention of ethical approval being sought. For the two
mixed methods studies, assessing the individual components separately may not have
captured the added value of integrating them as complementary.
Notwithstanding these reservations, the objectives of most studies were clear and they
offered some justification for the use of either quantitative or qualitative methods. The
specific methods chosen seemed appropriate to the research questions posed by the
included studies, and mixed quantitative and qualitative methods were utilized to offer a
wider frame for evaluation in two studies. Overall it is difficult to assess the value of the grey
literature studies, either individually or collectively as a body of evidence. The characteristics
of the included studies and their findings are presented in Appendix 3.
3.4 NARRATIVE SYNTHESIS OF RESULTS

This section is organized according to the research questions introduced earlier in the
review.
Relationship between recovery and relapse; relationship between relapse and default
or return default/episodes of default
Of the 24 studies included in this review, only six addressed the issue of relapse and/or
reported relapse rates (Bahwere et al., 2014; Ciliberto et al., 2005; Linneman et al., 2007;
Querubin, 2006; Taylor, 2002; UNICEF, 2012), and none addressed the relationship
between relapse and default or return default. This may be partly attributed to the relatively
short duration of most interventions and the fact that most of the studies did not include post-
intervention follow-up.
The first of the six studies that addressed relapse was a randomized controlled clinical
effectiveness trial conducted by Bahwere et al. (2014; high quality), comparing the
effectiveness of a whey protein-based RUTF with a standard peanut-based RUTF for
treating SAM in central Malawi. The article briefly mentioned that children included in the
intervention who then relapsed were not eligible to be readmitted to the programme,
although the rate of relapse was not reported. A second clinical effectiveness trial conducted
by Ciliberto et al. (2005; medium quality) compared the effectiveness of home-based therapy
using RUTFs with standard inpatient therapy using F-75 and F-100 for treating SAM in
southern Malawi. Recovery and relapse or death rates were reported as the primary
outcomes of the study. The authors found that children with SAM who received the home-
based therapy were more likely to achieve a WHZ > -2 compared with those who received
the standard therapy (79 percent compared with 46 percent; p < 0.001) and were less likely
to relapse or die (8.7 percent compared with 16.7 percent; p < 0.001). The authors reported
a default rate of 9.8 percent for children who received the home-based therapy and 8.1
percent for those who received the standard therapy. The authors did not identify specific
reasons for relapse or default among those children included in the intervention, nor did they
attempt to draw a relationship between relapse and default.
A prospective cohort study was conducted by Linneman et al. (2007; high quality) to study
the outcomes of home-based therapy with RUTF to treat both children with SAM and
children with MAM in an operational setting in southern Malawi. Of the 2,131 severely
malnourished children included in the intervention, an average of 89 percent recovered, 3
percent failed to recover or relapsed, 1 percent died and 7 percent defaulted across the 12
rural centres studied. Of the 806 moderately malnourished children included in the
intervention, 85 percent recovered, 4 percent failed to recover or relapsed, 2 percent died
and 8 percent defaulted on average across the 12 rural centres. The authors did not report
relapse rates separately from non-recovery rates or identify reasons for relapse among the
study population.
Three programme evaluation reports identified in the grey literature addressed the issue of
relapse. One of these reports evaluated an ACF-USA therapeutic feeding programme to
treat SAM in South Sudan (Querubin, 2006; low quality). All children included in the
intervention began treatment at a therapeutic feeding centre (TFC) utilizing F-75 for the first
phase of their treatment; they either continued to receive inpatient care utilizing F-100 for the
second phase of the programme (28 percent) or transitioned to home treatment utilizing
RUTF (72 percent). Of those children in the home treatment group, 81 percent recovered
and 7 percent defaulted. The author reported that there were no cases of relapse among the
children in the home treatment group. A second programme report evaluated an outpatient
therapeutic feeding programme (OTP) to treat SAM and MAM in 10 rural council areas in
North Darfur, Sudan (Taylor, 2002; low quality). Children with SAM included in the
intervention were treated with RUTF and children with MAM were treated with RUSF over a
five-month period in 2001. A supplementary feeding programme (SFP) targeting children
and pregnant and lactating women was also established. The average rate of discharge from
the OTP to the SFP was 81.4 percent, the average rate of default was 10.1 percent, the
average rate of mortality was 2.9 percent and the average rate of transfer to a TFC, hospital
or dispensary was 5.6 percent across the 10 rural councils. The authors reported that
readmission rates were approximately 1 percent of total admissions.
The third programme report evaluated four components of an integrated management of

acute malnutrition (IMAM) programme in Kenya, including community outreach, outpatient
treatment for SAM cases without medical complications, inpatient treatment for SAM cases
with medical complications and management of MAM (UNICEF, 2012; low quality). UNICEF
reported that outpatient SAM treatment utilizing RUTF resulted in a recovery rate of 80.7
percent, a mortality rate of 1.5 percent, a default rate of 12.9 percent and a relapse rate of
3.2 percent. Inpatient SAM treatment for children with medical complications utilizing F-75
and F-100 resulted in a recovery rate of 84.6 percent, a mortality rate of 8.7 percent, a
default rate of 1.4 percent and a relapse rate of 6.1 percent. Treatment for children with
MAM utilizing supplementary foods resulted in a recovery rate of 80.5 percent, a mortality
rate of 0.4 percent, a default rate of 14.4 percent and a relapse rate of 3.7 percent (note:
these rates were averaged across eight districts).
Reasons for default and relapse or return defaults/episodes of default
All 24 of the studies included in this review provided data about rates of default. Reasons for
default were often cited, including distance to travel to sites, family illness and other
commitments, other household priorities and poor community sensitization and household
follow-up. As related above, few studies reported relapse rates and only one identified
possible reasons for relapse within the context of the study. The UNICEF programme
evaluation report (2012; low quality) suggested that the lack of a follow-up system to track
children treated for SAM or MAM and encourage them to return for outpatient treatment may
have contributed to the relapse rate, as well as the sharing of RUTFs among siblings and
other non-admitted children. Improved community outreach was identified as a mechanism
for addressing these issues.
A mixed methods study of an OTP for SAM treatment in Ethiopia conducted by Belachew
and Nekatibeb (2007; medium quality) identified several reasons for a 44.8 percent default
rate in one administrative region (Oromia), including a lack of supplies, attrition of trained
staff and poor community mobilization, all of which caused interruptions in service. The
authors also attributed the default rate to the long distances that mothers and children had to
travel for follow-up services, and recommended that two weeks worth of RUTF be provided
at a time and that the programme be decentralized to community health posts in order to
reduce opportunity costs for participants and their families. Another mixed methods study
comparing maternal attitudes towards supplementary foods, specifically CSBs and lipid-
based nutrient supplements (LNSs), to treat MAM in Malawi found that, while mothers had
positive attitudes towards both, they were likely to stop giving LNS if children were sick
because the LNS was considered sweet in taste (Flax et al., 2009; medium quality).
Of the observational cohort studies included in the review (n=7), two offered reasons for
default. Lagrone et al. (2010; medium quality) studied the effectiveness of using a locally
produced soy/peanut RUSF to treat MAM in southern Malawi and suggested that a relatively
low default rate (4 percent) compared with similar interventions utilizing CSB may be due to
lower opportunity costs incurred by mothers, who had to gather firewood and prepare food in
order to utilize CSB. Linneman et al. (2007; high quality, observational cohort study)
indicated that high default rates at two rural centres located near border areas along a major
truck route may have been caused in part by undiagnosed HIV infection among children in
these centres. The authors suggested that routine HIV testing and home health visits by
village health workers might be added to interventions in these communities.
A programme evaluation report from an outpatient SAM treatment intervention conducted in

Afghanistan identified four possible reasons for a relatively high default rate of 23.8 percent,
based on interviews with mothers: illness of the child or mother; children accompanying
mothers to cultivate land; insufficient food distributed at the centres; and problems at home
(Lurqin, 2003; low quality). Taylor (2002; low quality, programme evaluation report) attributed
a high default rate of 34 percent in one rural council area in Darfur to children from
pastoralist families being in town for only a few days at a time. A programme evaluation
report of an outpatient programme conducted in South Sudan identified TFCs being cut off
by heavy rains and mothers having to walk with food and a child for four hours as reasons
for a default rate of 15 percent (Walker, 2004; low quality). Default rates for outpatient
treatment in some districts involved in UNICEFs Kenya intervention were partly attributed to
weak household follow-up (UNICEF, 2012; low quality).
A randomized controlled clinical effectiveness trial conducted by Oakley et al. (2010; high
quality) compared a standard RUTF made with 25 percent milk powder with one made with a
10 percent milk powder/soy blend. The authors addressed the possibility of higher default
rates in the 10 percent milk powder RUTF group due to subtle differences in taste or colour,
but did not find sufficient evidence either to confirm this or rule it out (based on a 3 percent
default rate for both intervention groups).
4 CONCLUSION
This review identified 24 publications and programme reports that reported recovery, relapse
and default and reasons for recovery and default in the management of acute malnutrition in
humanitarian emergencies. Assessment of these publications did not provide sufficient
information for conclusions to be drawn in relation to the review questions.
The relationship between recovery and relapse; and between relapse and default or
return default/episodes of default: The evidence related to this theme is inconclusive.
Reasons for default and relapse or return defaults/episodes of default: The strength of
evidence relating to this theme is limited.
This review provided further confirmation that RUTF used as part of the WHO 2013 protocol
is effective in promoting recovery from SAM and reducing mortality. We could not establish
whether default rates reported were lower according to the WHO 2013 protocol. We also
found limited data related to relapse. Similarly, as reported in previous reviews, we found
that the use of RUSF to treat MAM, compared with other supplementary foods such as FBFs
or CSBs, has the potential to improve nutritional recovery for children suffering from MAM.
However, the weight gain reported by studies for children given RUSF or CSB was small.
There were no substantial differences in mortality rates reported for children who received
RUSF compared with children in CSB groups.
The review found very little evidence on relapse from programmes implemented to manage
MAM and SAM in emergencies. Due to the nature of humanitarian emergencies, it is difficult
to conduct longitudinal data collection and trace the same children once they have been
discharged, except in refugee camp settings. Low mortality and relapse rates were
associated with access to skilled medical care, availability of essential drugs, greater
parental awareness of the consequences of illness and the need to seek available health
services early.
Reasons for default were often cited and included travel distance to sites, family illness and
other commitments, other household priorities, and poor community sensitization and
household follow-up. However, none of the included studies focused on defaulted children
as an objective.
We found that the challenges to successful completion of acute malnutrition treatment were
multi-factorial and included geographic, cultural and socioeconomic obstacles to care
(Guerrero et al., 2013). For successful implementation of the WHO protocols for CMAM,
emergency nutrition workers need to be able to adapt protocols to different contexts.
Programmatic issues and operational challenges also vary between settings and are
powerful determinants of the efficacy of programmes. Additional factors were the degree of
integration into existing health systems, the extent of service coverage, staff training and
alignment of inpatient treatment with WHO guidelines. However, there were pervasive
problems such as resource constraints, the lack of adequate supervision and monitoring,
lack of adequate outreach to identify the most vulnerable children, dependence on external
technical support, weak pipelines for specialized nutritious foods and logistical constraints.
4.1 LIMITATIONS OF THE REVIEW

There is considerable heterogeneity in the evidence base owing to the diversity of study
types, types of intervention and the settings in which programmes are delivered. In addition,
the quality of the published literature varies markedly. Generally, the grey literature reports
and qualitative studies included in the review lacked crucial detail, such as details of their
recruitment strategies. Review articles did identify the presence of significant barriers
relating to the acceptability of the intervention and implementation issues. These qualitative
insights also highlight the importance of the context in which malnutrition treatment
programmes are implemented and the challenges encountered, as illustrated by the
variability of effect of seemingly similar interventions based on identical guidelines and
protocols. This suggests that interventions cannot simply be copied from one setting and
implemented in another, but that some local adaptation is required. The detail presented in
the mixed methods studies offers a more nuanced understanding of the precise context and,
indeed, suggests that some factors are generic and that others, although present, may not
have the same relative importance in a different context. The three main limitations of this
review are outlined below.
One major limitation is the lack of qualitative data, as only two mixed methods studies were
eligible for inclusion following peer review feedback (Belachew and Nekatibeb, 2007; Flax et
al., 2009). It was assumed during the protocol development that the search would yield more
qualitative studies, but this was not the case. Upon applying the stricter interpretation of
humanitarian emergencies, most of the 22 qualitative studies were judged not to derive from
humanitarian emergencies and therefore they were excluded from the review. However, it is
possible that qualitative studies related to treatment of acute malnutrition in humanitarian
emergencies are part of larger response documents, and therefore more difficult to find
given the search strings employed. Also, it was expected that practitioners in the field of
malnutrition in humanitarian emergencies would be able to provide studies to the review
team. While efforts were made to contact experts, the pool of experts was limited to only five
organizations Concern Worldwide, Action Contre La Faim (ACF), Mdecins Sans
Frontires (MSF), the International Committee of the Red Cross (ICRC) and KEMRI-
Wellcome Trust and two universities (Washington University in St. Louis, USA and the
Universit Catholique de Louvain in Belgium).
4.2 INFLUENCE OF OPERATIONAL REALITIES ON

PROGRAMMING AND RESEARCH
Few of the studies cited challenges relating to operational realities affecting programmes to
manage acute malnutrition. While context was given in relation to types of population (i.e.
pastoralists, other migrating populations), security and access were rarely mentioned. Few
details were provided about how studies were interrupted or affected by ongoing
humanitarian emergencies. For example, we would have expected to see more information
related to weaknesses in supply chain management in relation to specialized nutritious
foods. This highlights an important gap in the literature, since understanding of how
humanitarian emergencies affect the provision of services is important to document.
4.3 AGE AND SEX DATA

While child-level data disaggregated by sex and age is routinely collected in the
management of acute malnutrition programmes, this data was not reported in the included
studies. This creates a gap in understanding how the management of acute malnutrition may
result in default and/or relapse differently in boys versus girls. Given that boys are more
likely to be wasted than girls in humanitarian emergency settings (Wamani et al., 2007),
there could be a different type of programming modality for boys compared with girls at
certain ages. The lack of age breakdown also represents a gap in this review, since children
aged 623 months are often the most wasted age group (UNICEF, 2013). Age category
analysis can inform how programmes can adapt to age-specific diet and psychosocial needs
in order to mitigate relapse and default and improve recovery.
4.4 RECOMMENDATIONS FOR FUTURE RESEARCH

This review found few studies that explored the issue of relapse. In order to understand the
relationship between relapse and recovery, more research on default and relapse post-
intervention is needed. A first step would be to conduct secondary analysis on existing
programme data (individual-level databases on SAM and MAM performance are managed
by nutrition clusters/sectors during humanitarian emergencies) and identify which countries
have the highest levels of default and relapse. With this prioritization, specific geographic
areas could be further researched with follow-up studies.
Additionally, to build the evidence base, prospective research could focus on one country
with quality programme data and high default and/or high relapse rates and follow-up studies
could be conducted of children who have defaulted or relapsed. Such studies will allow for
an improved understanding of why treatment outcomes were not maintained after discharge.
This may have implications related to guideline modifications: dosage, minimum length of
stay; sanitation and hygiene situation in the household leading to repeated illnesses;
complementary food adequacy in the household; household sharing of specialized nutritious
foods; etc.
Future studies should also examine how engagement with community members and care-
givers through the implementation of nutrition education activities, counselling and home
visits contributes to successful programme delivery and improvement of intervention
outcomes, particularly default and relapse. High-quality longitudinal observational studies
are needed to investigate the long-term impact on children of treatment for acute
malnutrition. Post-discharge follow-ups are needed to assess long-term growth and
development differences, mortality and relapse.
There are studies ongoing which examine the linkages between wasting and stunting (Khara
and Dolan, 2014). These analyses examine the implications of a childs nutritional status, i.e.
being stunted, before experiencing an episode of wasting, and the long-term morbidity and
mortality effects of wasting.
APPENDIX 1: DATABASE SEARCH
STRATEGY STRINGS
INITIAL SEARCH STRATEGY/OUTPUTS OF SCOPING
SEARCH
Database: Ovid Medline In-process and other non-indexed citations and Ovid Medline
<1946 to Present>
Search strategy
# Searches
1 malnourish$ or undernourish$ or malnutrition or undernutrition or kwashiorkor or marasmus).ti,ab. (39065)
2 Malnutrition/ (7780)
3 exp Protein-Energy Malnutrition/ (8495)
4 1 or 2 or 3 (44198)
5 Child, Preschool/ or Infant/ (1023482)
6 child$ or baby or babies or infant$).ti,ab. (1300841)
7 5 or 6 (1772685)
8 (humanitarian$ or conflict$ or famine$ or disaster$ or emergency relief or sub-saharan africa or south-east
asia or latin america$).ti,ab. (130391)
9 Relief Work/ or Disaster Planning/ or Starvation/ (22503)
10 8 or 9 (146588)
11 Child Nutrition Disorders/ (2457)
12 10 and 11 (122)
13 4 and 7 (16526)
14 10 and 13 (813)
15 or 14 (851)
FINAL SEARCH STRATEGY

# Search strategy
1 (malnourish$ or undernourish$ or under-nutrition or under-nourish or mal-nutrition or malnutrition or
undernutrition or wasting or kwashiorkor or marasmus or SAM or severe-acute-malnutrition).ti,ab.
2 Malnutrition/
3 exp Protein-Energy Malnutrition/
4 1 or 2 or 3
5 Child, preschool/ or Infant/
6 (child$ or baby or babies or infant$ or pre-school or preschool).ti,ab.
7 5 or 6
8 4 and 7
9 Child Nutrition Disorders/ or Infant Nutrition Disorders/
10 8 or 9
11 (humanitarian$ or conflict$ or famine$ or disaster$ or emergency or emergencies or Developing countr$ or
sub-saharan africa or south-east asia or latin america$ or Ethiopia or Malawi or South Sudan or Niger or
Pakistan or India or Bangladesh or Kenya or Nepal or Afghanistan or Angola or Armenia or Azerbaijan or
Bosnia-Herzegovina or Burkina Faso or Burundi or Republic of Congo or Democratic Republic of Congo or
El Salvador or Eritrea or Gambia or Georgia or Guatemala or Haiti or Honduras or Indonesia or Iraq or
Korea DPR or Laos or Liberia or Mauritania or Mongolia or Mozambique or Nicaragua or Panama or Papua
New Guinea or Rwanda or Senegal or Sierra Leone or Somalia or Tajikistan or Tanzania or Uganda).ti,ab.
# Search strategy
12 Developing Countries/ or Africa South of the Sahara/ or Asia, Southeastern/ or Latin America/ or Ethiopia/
or Malawi/ or Sudan/ or Niger/ or Pakistan/ or India/ or Bangladesh/ or Kenya/ or Nepal/ or Afghanistan/ or
Angola/ or Armenia/ or Azerbaijan/ or Bosnia-Herzegovina/ or Burkina Faso/ or Burundi/ or Democratic
Republic of the Congo/ or Congo/ or El Salvador/ or Eritrea/ or Gambia/ or Georgia (Republic)/ or
Guatemala/ or Haiti/ or Honduras/ or Indonesia/ or Iraq/ or Democratic Peoples Republic of Korea/ or
Laos/ or Liberia/ or Mauritania/ or Mongolia/ or Mozambique/ or Nicaragua/ or Panama/ or Papua New
Guinea/ or Rwanda/ or Senegal/ or Sierra Leone/ or Somalia/ or Tajikistan/ or Tanzania/ or Uganda/
13 Relief Work/ or Disaster Planning/ or Starvation
14 11 or 12 or 13
15 ((community adj3 care) or (home adj3 care) or (therapeutic$ adj3 food$) or (therapeutic$ adj3 diet$) or
(enrich$ adj3 food$) or (enrich$ adj3 diet$) or therapy or management or feeding or fortified food$ or
Plumpynut or F-100 or F-75 or nutritional or rehabilitation or outpatient$ or dietary intervention$ or Ready-
to-Use Therapeutic Food or RUTF).ti,ab.
16 Food, Fortified/ or Food Services/ or Community Health Services/ or Food, Formulated/ or Dietary
Supplements/ or Outpatients/
17 15 or 16
18 10 and 14 and 17
GREY LITERATURE SEARCH

Action Against Hunger
Centre for Reviews and Dissemination (CRD)
Concern Worldwide
Coverage Monitoring Network
DFID R4D
ELDIS
Emergency Nutrition Network (ENN)
Enhancing Learning and Research for Humanitarian Assistance (ELRHA)
European Commission Humanitarian Aid and Civil Protection Department
EvidenceAid
Feinstein International Center and Friedman School of Nutrition Science and Policy, Tufts University
Food and Nutrition Bulletin
Free from Hunger
Global Alliance for Improved Nutrition
Harvard Humanitarian Initiative (HHI)
Health Technology Assessment Group
Helen Keller International
Humanitarian Policy Group (HPG)
Humanitarian Practice Network
Humanitarian Social Network
International Association of Professionals in Humanitarian Assistance and Protection (PHAP)
International Initiative for Impact Evaluation (3ie)
International Medical Corps
Mdecins Sans Frontires (MSF)
Nutritional Causal Analysis
Overseas Development Institute (ODI)
Oxfam Policy Development and Evaluation Service
ReliefWeb
Save the Children
The CMAM Forum
The Network on Humanitarian Actions
The World Bank
UNICEF
USAID
Valid International
WHO Global Health Library
World Food Programme
APPENDIX 2: ASSESSMENT OF
INCLUDED STUDIES
QUALITY ASSESSMENT OF INCLUDED QUANTITATIVE CASE
CONTROL AND OBSERVATIONAL STUDIES AND GREY
LITERATURE
Exposure accurately measured
Rates or proportions reported
Provides implications for

Fits with other evidence
Recruitment acceptable
Can results be applied?

Follow-up long enough
Confounders identified
Clearly focused issues
Outcomes accurately
Results believable
Results precise
measured
practice
Grade
Publication
Amthor et al. (2009) Y Y Y Y N Y Y Y Y Y Y Y H
Chaiken et al. (2006) Y ? Y ? N ? Y ? ? ? Y Y L
Collins and Sadler (2002) Y Y Y Y N Y Y Y ? ? Y Y M
Deconinck (2004) Y ? ? N N N N ? ? ? N Y L
Gaboulaud et al. (2006) Y N Y ? N ? Y ? ? N ? Y L
Grellety et al. (2012) Y Y Y Y Y Y Y Y Y Y Y Y H
Linneman et al. (2007) Y Y Y Y Y Y Y Y Y Y Y Y H
Lagrone et al. (2010) Y Y Y Y ? Y Y ? ? ? Y Y M
Lurquin (2003) Y ? N N N N Y N N N Y N L
Morgan et al. (2015) Y ? N N N N Y N N N Y N L
Querubin (2006) Y Y N N N N Y ? ? N Y N L
Taylor (2002) Y ? ? N N N Y ? ? N Y N L
UNICEF (2012) Y Y ? ? N ? Y ? ? N Y Y L
Walker (2004) Y ? ? ? N ? Y ? ? N Y N L
Abbreviations and symbols: Y= Yes, N = No, ? = Not sure/unclear, H = High, M = Medium, L = Low
QUALITY ASSESSMENT OF INCLUDED QUANTITATIVE RCTS AND

GREY LITERATURE
Patients/study personnel/health
Groups similar at start of trial
Clinically important outcome
Benefits worth harms/cost

Can results be applied?
Clearly focused issues
workers blinded
Results precise
Treated equally
Randomized
Grade
Publication
Bahwere et al. (2014) Y Y Y Y Y Y Y Y Y H
Ciliberto et al. (2005) Y N N Y Y ? ? Y Y M
Huybregts et al. (2012) Y Y N Y ? ? ? Y Y M
Isanaka et al. (2009) Y Y N Y Y ? ? Y Y M
Karakochuk et al. (2012) Y Y N Y Y Y Y ? Y
Matilsky et al. (2009) Y Y Y Y Y Y Y Y Y H
Maust et al. (2015) Y Y N Y Y ? ? Y Y H
Oakley et al. (2010) Y Y Y Y Y Y Y Y Y M
Abbreviations and symbols: Y = Yes, N = No, ? = Not sure/unclear, H = High, M = Medium, L = Low
QUALITY ASSESSMENT OF INCLUDED QUALITATIVE STUDIES

AND GREY LITERATURE
Appropriate research design
researcher and participants

Appropriate data collection
Ethical issues considered

Qualitative methodology
Appropriate recruitment
Clear statement of aims
Data analysis rigorous

Relationship between
Value of research
Findings clear
appropriate
strategy
Grade
Publication
Belachew and Nekatibeb (2007) Y Y Y ? Y ? ? ? Y ? M
Flax et al. (2009) Y ? ? ? Y ? Y Y ? Y M
Abbreviations and symbols: Y = Yes, N = No, ? = Not sure/Unclear, H = High, M = Medium, L = Low
APPENDIX 3: INCLUDED PUBLICATIONS

REPORTED RESULTS
CHARACTERISTICS OF INCLUDED MIXED METHODS AND
OBSERVATIONAL COHORT STUDIES
Author/year Country/study Study Description of intervention Main findings
population design
Belachew Ethiopia, Mixed The intervention was community- The proportion of children who
and children 659 methods based management of children with achieved the recovery target (33
Nekatibeb months old with study SAM using RUTF in three districts. percent) at the end of 13 weeks was
(2007) SAM (n=355) All children received RUTF and were low; 44 percent defaulted and 0.8
(journal plus 36 key followed weekly until recovery, death percent died.
article) informants and or default. Duration of project was 5
30 focus groups 13 weeks.
Flax et al. Malawi Mixed This research studied attitudes of Both supplements were said to be
(2009) methods 504 Malawian mothers to LNS and highly acceptable, children learned to
(journal study CSB via exit interviews completed at eat them within two weeks, and
article) the end of three 12-week clinical mothers were willing to use them
trials, and compared differences again.
between the groups.
Amthor et al. Malawi, children Observational Each child received 200kcal of The proportion of children who
(2009) 660 months old cohort study RUTF. The children were followed recovered was 93.7 percent, and 1.8
(journal with SAM weekly until recovery, default or percent remained malnourished.
article) (n=826) death. Total duration of treatment Default and death rates were 3.6
was eight weeks or earlier if they met percent and 0.9 percent respectively.
the discharge criteria (attainment of Mean (SD) rate of weight gain was
WHZ >100 percent). 2.7 3.7g/kg/d. The proportion of
severe malnutrition before and after
was significantly less (p < 0.001).
Chaiken et al. Ethiopia, Observational All children received a daily dose of An 87 percent recovery rate was
(2006) children 659 cohort study 200kcal/kg/d RUTF per their body reported; 2.3 percent defaulted and
(journal months old with weight. They were monitored weekly fewer than 1 percent died. Under-five
article) SAM (n=5,799) till eight weeks or discharge if early mortality and SAM before the
recovery, or death. In addition CSBs intervention were 1.47/10,000 and
were provided to minimize sharing 1.0 percent; after the intervention
and as a supplement to RUTF at these rates improved to 0.45/10,000
home. and 0.6 percent (95 percent CI, 0.2
0.9).
Collins and Ethiopia, Observational WHO endorsed interventions to Eighty-five percent recovered, 4
Sadler (2002) children 6120 cohort study manage SAM: a daily dose of percent died and the default rate was
(journal months old with RUTFs, providing 733kJ/kg/d per 5 percent. Median rate of weight gain
article) SAM (n=170) body weight. Followed up until was 3.2g/kg/d (95 percent CI, 1.9
recovery or death or default. At 5.6).
home, Famix + RUTF given in
addition to Vitamin A (100,000 IU for
children <12 months and 200,000 IU
if older) plus 5mg folic acid
fortnightly.
Gaboulaud et Niger, children Observational A total of 660 children from the three Defaults from treatment were
al. (2006) 659 months old cohort study cohorts received care exclusively in reportedly high: 28.1 percent, 16.8
(journal with SAM hospital (inpatient care). 937 children percent and 5.6 percent respectively
article) (n=1,937) were initially treated at TFCs and for TFC only, TFC + home and home
then completed treatment at home, only. The overall case fatality for the
and 340 children were exclusively entire programme was 6.8 percent.
treated at home. Case fatality rates were higher in
TFC only (18.9 percent) compared
with TFC + home (1.7 percent) and
home only (0 percent).

population design
Grellety et al. Niger, children Observational Children were registered in the The absolute rate of wasting was
(2012) 623 months old cohort study programme to receive monthly 4.71 events per child-year in the
(journal with MAM distributions of RUSF (PlumpyDoz). intervention group and 4.98 events
article) (n=2,238) per child-year in the comparison
group. The intervention group had a
small but higher weight-for-length Z-
score gain (-0.2z vs. -0.3z) and less
loss of MUAC than the comparison
group (-2.8 vs. -4.0mm). Mortality
was lower for children whose
households received the intervention
than those who did not (adjusted HR
0.55, 95 percent CI, 0.320.98).
Lagrone et al. Malawi, children Observational Each child received 65kcal/kg/d of RUTF (milk-peanut and soy-peanut
(2010) 659 months old cohort study locally produced soy/peanut RUSF. fortified spread (FS)) resulted in
(journal with MAM Anthropometry measurements were greater weight than CSB (p = 0.01).
article) (n=2,417) taken every two weeks and additional Children who received either milk-
RUSFs were distributed at this time if peanut FS or soy-peanut FS also
the child remained wasted. Study gained weight faster than children
duration: eight weeks. receiving CSB. Eight percent of
children developed oedema.
Linneman et Malawi, children Observational Outcomes of severely malnourished Recovery rate in the severely
al. (2007) 660 months old cohort study children were compared with a malnourished group was 89 percent
(journal with SAM moderately malnourished group vs. 85 percent in the moderately
article) (n=2,131) and (control). Each child received a daily malnourished group. The death rate
MAM (n=806) dose of RUTF, providing 733kJ/kg/d for severely malnourished children
energy, and was monitored for eight was 1 percent vs. 2 percent for the
weeks or until early recovery or moderately malnourished (p < 0.01).
death. RWG was 3.5 4.1 (severe) vs. 4.6
4.1 (moderate) g/kg/day. Default
rate in the severely malnourished
group was 7 percent vs. 8 percent in
the moderately malnourished group.
CHARACTERISTICS OF INCLUDED GREY LITERATURE

PROGRAMME EVALUATION REPORTS
population design
Deconinck Ethiopia, children Programme Children enrolled with SAM Programme treating SAM achieved 86.6
(2004) (grey 659 months old evaluation received RUTF. Children percent recovery rate. Defaulting and
literature) with SAM report enrolled with MAM received mortality rates were 4 percent and 0.6 percent
(n=1,470) and supplementary foods in line with respectively. Outcomes for children treated
MAM (n=5,558) WHO recommendations. for MAM were not reported.
Lurqin (2003) Afghanistan, Programme Each SAM child was provided Proportions of children who died in the
(grey children 659 evaluation with an average of 200kcal of programme or defaulted were 6 percent and
literature) months old with report RUTF PlumpyNut and BP100, 23 percent respectively. The proportion that
SAM (n=635) supplied weekly. They were achieved nutritional recovery was not
monitored until recovery, death reported. The rate of weight gain reported
or default or discharge at the end was 6.1g/kg/day.
of the programme.
Morgan et al. Angola, children Programme Children were treated based on The programme had a high cure rate. The
(2015) (grey 659 months old evaluation WHO guidelines. Children with SAM cure rate was 93.8 percent, and the
literature) with SAM report SAM received RUTF based on MAM cure rate 96.7 percent. Mortality and
(n=23,865) and their body weight. Children with default rates were 4.8 percent and 1.0
MAM (n=53,229) MAM received RUSF each day. percent respectively for SAM, and 2.9 percent
and 0.2 percent respectively for MAM.
Querubin Sudan, children Programme Children received systematic Average recovery rate reported was 81
(2006) (grey 659 months old evaluation medical treatment, plus F-75 percent. Average weight gain was 9g/kg/day.
literature) with SAM report therapeutic milk, which provided Overall defaulting from TFC was low at 12
135ml/kg body weight/day in percent.
Phase 1. In Phase 2, children
received F-100 therapeutic milk,
which provided 200kcal/kg/day,
plus porridge. Duration of
treatment: four weeks.
Taylor (2002) Sudan, children Programme SAM cases with medical Recovery rate for children treated for SAM
(grey 659 months old evaluation complication were first treated in was 81.4 percent. The proportions of children
literature) with SAM report hospital stabilization centres who died while on treatment and those who
(n=836) and using F-75 plus medication. They defaulted from treatment were 2.9 percent
MAM (n=24,000) were transferred to outpatient and 10.1 percent respectively. Readmission
care and received RUTF when rate or relapse was 1.0 percent.
complications were resolved. Supplementary feeding outcomes were not
Children enrolled with MAM reported.
received RUSF.
UNICEF Kenya, children Programme Children received standard The high default rates were explained partly
(2012) (grey 659 months old evaluation treatment: 1) F-75 and F-100 at by weak household follow-up, whereas the
literature) with SAM report inpatient care, and RUTF at greater length of stay was often due to social
(n=18,509) and outpatient care for those pressures from parents and community
MAM (n=79,183) admitted with SAM, and 2) members who wanted children to benefit from
supplementary foods endorsed RUTF for as long as possible. Inadequate
by WHO for the treatment of tracking of weight gain was an additional
MAM. reason. Inpatient treatment of SAM achieved
a recovery rate of 84.6 percent, a death rate
of 8.7 percent, a default rate of 1.4 percent
and an overall relapse rate of 6.1 percent.
Outpatient treatment of SAM achieved an
80.7 percent recovery rate, 1.5 percent death
rate and 13 percent default rate in two
districts sampled. Default rates were high in
Kisumu district (17.3 percent) and Nairobi
district (19.3 percent). The average relapse
rate was low (3.2 percent). Interventions
delivered to manage MAM achieved an 80.5
percent recovery rate, 0.4 percent death rate
and 14.5 percent default rate. Relapse was
not reported for MAM.
Walker (2004) Sudan, children Programme The children were treated in line Stabilization centre management of SAM
(grey 659 months old evaluation with WHO recommendations. achieved 93.5 percent recovery, 2.6 percent
literature) with SAM report Children with medical conditions mortality and 0 percent default rates. In the
(n=726) were treated at hospital outpatient programme phase, the proportions
stabilization centres while of children who recovered, died and defaulted
children with no such were 81.8 percent, 3.1 percent and 15
complications were admitted percent respectively.
directly into outpatient treatment
programmes.
CHARACTERISTICS OF INCLUDED RANDOMIZED CONTROLLED

TRIALS
population design
Bahwere et Malawi, Randomized The children were evenly randomized The proportions of children who
al. (2014) children 659 controlled to either WPC-RUTF (n=292 in recovered, died and/or defaulted in the
(journal months old with trial intervention group) or P-RUTF (n=308 intervention group (WPC-RUTF) were
article) SAM (n=600) in an active control group). The 84.8 percent, 1.6 percent and 12.2
children in both groups were treated percent respectively. For children in
according to Malawian national the controlled treatment group (P-
guidelines for the management of RUTF) the recovery, mortality and
SAM. Children with medical default rates were 84.2 percent, 0.7
complications were referred to hospital percent and 12.2 percent respectively.
stabilization centres for inpatient care, The average rate of weight gain was
and were only eligible when they were low at 3.1g/kg/d; 4g/kg/day (in the
referred back. intervention group) and 2.9 g/kg/day
(in the control group).
Ciliberto et al. Malawi, Randomized The children were systematically The average recovery rate was 83
(2005) children 659 controlled allocated to either standard treatment percent, the death rate was 5 percent
(journal months old with trial (n=186) or home-based therapy with and drop-out was 4 percent. The
article) SAM (n=1,178) RUTF (n=992). children who received home-based
therapy with RUTF were more likely to
achieve a WHZ > -2 than those who
received standard therapy (93 percent
recovery rate compared with 46
percent; p < 0.001). The overall
relapse rate was 8.2 percent. Children
in the RUTF group were also less
likely to relapse (3.5 compared with
2.0g/kg/d).
Huybregts et Chad, children Randomized Intervention group received a monthly Adding RUSF to a package of monthly
al. (2012) 636 months controlled quantity of RUSF (PlumpyDoz), household food rations for households
(journal with MAM trial representing a daily dose of 46g containing a child assigned to the
article) (n=1,038) (~247kcal/d). All participating care- intervention group did not result in a
givers in both groups received a reduction in cumulative incidence of
monthly food package. Study duration: wasting (incidence risk ratio: 0.86; 95
four months. percent CI: 0.67, 1.11; p = 0.25).
However, the intervention group had a
modestly higher gain in height-for-age
(+0.03 Z-score/mo; 95 percent CI:
0.01, 0.04; p < 0.001).
Isanaka et al. Niger, children Randomized One packet per day of RUTF Reduction in wasting and severe
(2009) 660 months controlled (PlumpyNut) versus no intervention. wasting; impact greater during the
(journal old with MAM trial period of documented food shortage;
article) (n=3,533) no effect on mortality.
Karakochuk Ethiopia, Randomized Children were randomly assigned to The recovery rate was higher in the
et al. (2012) children 660 controlled one of two foods. Children received RUSF group (73 percent) than in the
(journal months old with trial CSB or RUSF bi-weekly. Study CSB group (67 percent), but no
article) MAM (n=1,125) duration: 16 weeks. significant difference was found (p =
0.056). The overall HR for CSB was
0.85 (95 percent CI, 0.730.99), which
indicates that the CSB group had a 15
percent lower overall recovery rate (p
= 0.039). Defaulting and death rates
were low and non-response rates
were high. Neither group met the
Sphere standards for recovery.
Matilsky et al. Malawi, Randomized Children were assigned to one of three Each RUSF (milk-peanut and soy-
(2009) children 660 controlled diets: A) milk/peanut FS, B) peanut FS) resulted in greater weight
(journal months old with trial soy/peanut FS (RUTF) and C) gain than CSB (p = 0.01). Children
article) MAM (n=1,326) corn/soy blends (CSBs). They were who received either milk-peanut FS or
assessed weekly until recovery or for soy-peanut FS gained weight faster
the eight-week duration of the than children receiving CSB. Eight
programme. percent of children developed
oedema.

population design
Maust et al. Sierra Leone, Randomized Children were assigned to two Average recovery rate in the
(2015) children 659 controlled treatment groups: 1) integrated integrated management group
(journal months old with trial management of MAM and SAM; 2) (including MAM and SAM children)
article) SAM (n=857) standard management of MAM and was 83 percent compared with 79
and MAM SAM. In 1) children with SAM received percent. Proportions of children who
(n=1,100) RUTF plus amoxicillin at 175 died while receiving treatment were 4
kcal/kg/d, while children with MAM percent for both integrated
received only RUTF at 175 kcal/kg/d. management and standard
In 2) SAM children received RUTF management groups. The default rate
(120kcal/kg/d) plus medication, while was 8 percent for both groups. Rates
MAM children received FBFs. Care- of weight gain during the first four
givers received nutrition information to weeks of treatment were 3.0
prevent malnutrition. SAM follow-up 0.2g/kg/d and 3.8 0.2g/kg/d.
was weekly until recovery or end of
intervention. MAM follow-up was every
14 days. There was no limit to the
duration of the intervention.
Oakley et al. Malawi, Randomized Random allocation of children to The recovery rate among children
(2010) children 659 controlled receive two locally produced RUTFs, receiving 25 percent milk RUTF was
(journal months old with trial one containing 25 percent milk greater than for children receiving 10
article) SAM (n=1,874) (n=945) and the other containing 10 percent milk RUTF, starting from 64
percent milk (n=929). The children percent compared with 57 percent
were followed until the end of the after four weeks, and 85 percent
project (eight weeks) or early compared with 81 percent after eight
recovery, death or default. weeks (p < 0.18). Children receiving
25 percent milk RUTF also had a
higher rate of weight gain (2.4
2.8g/kg/day) compared with those
receiving 10 percent milk RUTF (1.9
2.7g/kg/day) (p < 0.001).
REFERENCES
INCLUDED PUBLICATIONS
Amthor, R.E., Cole, S.M. and Manary, M.J. (2009). The use of home-based therapy with
ready-to-use therapeutic food to treat malnutrition in a rural area during a food crisis. Journal
of the American Dietetic Association, 109(3), 464-467.
Bahwere, P., Banda, T. et al. (2014). Effectiveness of milk whey protein-based ready-to-use
therapeutic food in treatment of severe acute malnutrition in Malawian under-5 children: a
randomised, double-blind, controlled non-inferiority clinical trial. Maternal & Child Nutrition,
10(3), 436-451.
Belachew, T. and Nekatibeb, H. (2007). Assessment of outpatient therapeutic programme

for severe acute malnutrition in three regions of Ethiopia. East African Medical Journal,
84(12), 577.
Chaiken, M.S., Deconinck, H. and Degefie, T. (2006). The promise of a community-based

approach to managing severe malnutrition: A case study from Ethiopia. Food and Nutrition
Bulletin, 27(2), 95-104.
Ciliberto, M.A., Sandige, H. et al. (2005). Comparison of home-based therapy with ready-to-
use therapeutic food with standard therapy in the treatment of malnourished Malawian
children: a controlled, clinical effectiveness trial. The American Journal of Clinical Nutrition,
81(4), 864-870.
Collins, S. and Sadler, K. (2002). Outpatient care for severely malnourished children in
emergency relief programmes: a retrospective cohort study. The Lancet, 360(9348), 1824-
1830.
Deconinck, H. (2004). Adopting CTC from Scratch in Ethiopia (Special Supplement 2).
Supplement 2: Community-based Therapeutic Care (CTC), November 2004, p.30.
www.ennonline.net/fex/102/4-2
Flax, V. L., Thakwalakwa, C. et al. (2009). Malawian mothers attitudes towards the use of
two supplementary foods for moderately malnourished children. Appetite, 53(2), 195-202.
Gaboulaud, V., Dan-Bouzoua, N. et al. (2007). Could nutritional rehabilitation at home

complement or replace centre-based therapeutic feeding programmes for severe
malnutrition?. Journal of Tropical Pediatrics, 53(1), 49-51.
Grellety, E., Shepherd, S. et al. (2012). Effect of mass supplementation with ready-to-use
supplementary food during an anticipated nutritional emergency. PLoS One, 7(9), e44549.
Huybregts, L., Houngb, F. et al. (2012). The effect of adding ready-to-use supplementary
food to a general food distribution on child nutritional status and morbidity: a cluster-
randomized controlled trial. PLoS Med, 9(9), e1001313.
Isanaka, S., Nombela, N. et al. (2009). Effect of preventive supplementation with ready-to-
use therapeutic food on the nutritional status, mortality, and morbidity of children aged 6 to
60 months in Niger: a cluster randomized trial. JAMA, 301(3), 277-285.
Karakochuk, C., van den Briel, T., Stephens, D. and Zlotkin, S. (2012). Treatment of
moderate acute malnutrition with ready-to-use supplementary food results in higher overall
recovery rates compared with a corn-soya blend in children in southern Ethiopia: an
operations research trial. The American Journal of Clinical Nutrition, 96(4), 911-916.
Lagrone, L., Cole, S. et al. (2010). Locally produced ready-to-use supplementary food is an
effective treatment of moderate acute malnutrition in an operational setting. Annals of
Tropical Paediatrics, 30(2), 103-108.
Linneman, Z., Matilsky, D. et al. (2007). A large-scale operational study of home-based

therapy with ready-to-use therapeutic food in childhood malnutrition in Malawi. Maternal &
Child Nutrition, 3(3), 206-215.
Lurquin, E. (2003). Ambulatory treatment of severe malnutrition in Afghanistan. Field

Exchange 19, July 2003, p.13. www.ennonline.net/fex/19/ambulatory
Matilsky, D.K., Maleta, K., Castleman, T. and Manary, M.J. (2009). Supplementary feeding
with fortified spreads results in higher recovery rates than with a corn/soy blend in
moderately wasted children. The Journal of Nutrition, 139(4), 773-778.
Maust, A., Koroma, A.S. et al. (2015). Severe and Moderate Acute Malnutrition Can Be
Successfully Managed with an Integrated Protocol in Sierra Leone. The Journal of Nutrition,
145(11), 2604-2609.
Morgan, S., Bulten, R. and Jalipa, H. (2015). Community case management approach to
SAM treatment in Angola. Field Exchange 49, March 2015, p.3.
www.ennonline.net/fex/49/angola
Oakley, E., Reinking, J. et al. (2010). A ready-to-use therapeutic food containing 10% milk is
less effective than one with 25% milk in the treatment of severely malnourished children. The
Journal of Nutrition, 140(12), 2248-2252.
Querubin, J. (2006). Home treatment for severe malnutrition in South Sudan. Field
Exchange 28, July 2006, p.17. www.ennonline.net/fex/28/southsudan
Taylor, A. (2002). Outpatient therapeutic programme (OTP): an evaluation of a new SC UK

venture in North Darfur, Sudan (2001). Field Exchange 16, August 2002, p.25.
www.ennonline.net/fex/16/outpatient
UNICEF (2012). Evaluation of integrated management of acute malnutrition IMAM: Kenya

country case study. New York, UNICEF.
Walker, E. (2004). CTC from Scratch Tear Fund in South Sudan (Special Supplement 2).
Supplement 2: Community-based Therapeutic Care (CTC), November 2004, p.28.
www.ennonline.net/fex/102/4-1
OTHER PUBLICATIONS
Aguayo, V.M., Agarwal, V. et al. (2013). Integrated programme achieves good survival but
moderate recovery rates among children with severe acute malnutrition in India. The
American Journal of Clinical Nutrition, 98(5), 1335-1342.
Akparibo, R., Lee, A. et al. (2016). Relationships between recovery and relapse, and default
and repeated episodes of default in the management of acute malnutrition in children in
humanitarian emergencies: A systematic review protocol.
Alcoba G., Kerac, M. et al. 2013. Do Children with Uncomplicated Severe Acute Malnutrition
Need Antibiotics? A Systematic Review and Meta-Analysis. PLoS ONE 8 (1): E53184.
Ashworth, A. (2006). Efficacy and effectiveness of community-based treatment of severe

malnutrition. Food and Nutrition Bulletin, 27(3 suppl3), S24-S48.
Bagchi, K. et al. (2004). Nutrition in humanitarian crises. Eastern Mediterranean Health

Journal, Vol. 10, No. 6, 2004.
Bhutta, Z.A., Ahmed, T. et al. (2008). What works? Interventions for maternal and child
undernutrition and survival. The Lancet, 371(9610), 417-440.
Booth, A., Harris, J. et al. (2013). Towards a Methodology for Cluster Searching to Provide
Conceptual and Contextual Richness for Systematic Reviews of Complex Interventions:
Case Study (Cluster). BMC Medical Research Methodology, 13(1), 118.
Briend, A., Lacsala, R. et al. (1999). Ready-to-use therapeutic food for treatment of
marasmus. The Lancet, 6736(99), 01078-8.
Brundtland, G.H. (2000). Nutrition and infection: malnutrition and mortality in public health.
Nutrition Reviews, 58(2), S1.
Burza, S., Mahajan, R. et al. (2015). Seasonal effect and long-term nutritional status
following exit from a Community-Based Management of Severe Acute Malnutrition program
in Bihar, India. European Journal of Clinical Nutrition.
Collins, S. (2004). Community-based therapeutic care. A new paradigm for selective feeding
in nutritional crises. Humanitarian Practice Network. London: Overseas Development
Institute.
Collins, S., Dent, N. et al. (2006). Management of severe acute malnutrition in children. The
Lancet, 368(9551), 1992-2000.
Critical Appraisal Skills Programme (2013). CASP Qualitative Checklist: 10 ways to help you
make sense of qualitative research. Critical Appraisal Skills Programme: Oxford, UK.
http://media.wix.com/ugd/dded87_29c5b002d99342f788c6ac670e49f274.pdf
Dale, N.M., Myatt, M., Prudhon, C. and Briend, A. (2013). Using mid-upper arm
circumference to end treatment of severe acute malnutrition leads to higher weight gains in
the most malnourished children. PLoS ONE, 8(2), e55404.
Diop, E.I., Dossou, N. et al. (2004). Home-based rehabilitation for severely malnourished
children using locally made ready-to-use therapeutic food (RUTF). In Pediatric
Gastroenterology, Hepatology and Nutrition: Second World Congress, Paris, France.
Force, M.T. (2012). Moderate Acute Malnutrition: A decision tool for emergencies. Geneva:
IASC Global Nutrition Cluster.
Gera, T. (2010). Efficacy and safety of therapeutic nutrition products for home based
therapeutic nutrition for severe acute malnutrition: a systematic review. Indian Pediatrics,
47(8), 709-718.
Giugliani, C., Duncan, B.B. et al. (2010) The impact of a short-term intervention using the
WHO guidelines for the management of severe malnutrition at rural facility in Angola.
Archives of Disease in Childhood 95:198-202.
Global Nutrition Cluster MAM Taskforce (2014). Moderate Acute Malnutrition a decision tool
for emergencies. http://nutritioncluster.net/wp-content/uploads/sites/4/2014/07/MAM-
Decision-Tool-June-2014.pdf
Guerrero, S. and Rogers, E. (2013). Access for All, Volume 1: Is community-based treatment
of severe acute malnutrition (SAM) at scale capable of meeting global needs?. Coverage
Monitoring Network, London. http://www.coverage-monitoring.org/wp-
content/uploads/2013/07/AAH-Policy-Paper-Comp.pdf
Kerac, M., Bunin, J. et al. (2009). Probiotics and Prebiotics for Severe Acute Malnutrition
(PRONUT Study): A Double-Blind Efficacy Randomised Controlled Trial in Malawi. The
Lancet, 374(9684), 136-44.
Khara, T. and Dolan, C. (2014). Technical Briefing Paper. The relationship between wasting
and stunting: policy, programming and research implications.
Lenters, L.M., Wazny, K. et al. (2013). Treatment of severe and moderate acute malnutrition
in low-and middle-income settings: a systematic review, meta-analysis and Delphi process.
BMC Public Health, 13(3), 1.
Manary, M.J. et al. (2004). Home based therapy for severe malnutrition with ready-to-use
food. Archives of Disease in Childhood 89.6: 557-561.
Myatt, M., Khara, T. and Collins, S. (2006). A review of methods to detect cases of severely
malnourished children in the community for their admission into community-based
therapeutic care programs. Food and Nutrition Bulletin, Vol. 27, Issue 3_suppl3, pp s7-s23.
Ndekha, M.J., Manary, M.J., Ashorn, P. and Briend, A. (2005). Home-based therapy with
ready-to-use therapeutic food is of benefit to malnourished, HIV-infected Malawian children.
Acta Paediatrica, 94(2), 222-225.
Nutriset (2008). PlumpyNut: un bonbon thrapeutique. Cahiers Agricultures 17.3: 327-327.
PAHO/WHO (2003). Guiding Principles for Complementary Feeding of the Breastfed Child.
Washington, DC.
http://www.who.int/nutrition/publications/guiding_principles_compfeeding_breastfed.pdf
Picot J., Hartwell D. et al. (2012). The Effectiveness of Interventions to Treat Severe Acute
Malnutrition in Young Children: A Systematic Review. Health Technology Assessment 16
(19): 1315.
Puett, C. and Guerrero, S. (2015). Barriers to access for severe acute malnutrition treatment
services in Pakistan and Ethiopia: A comparative qualitative analysis. Public Health Nutrition,
18(10), 1873-1882.
Puett C., Hauenstein Swan, S. and Guerrero, S. (2013). Access for All, Volume 2: What
factors influence access to community-based treatment of severe acute malnutrition?.
Coverage Monitoring Network.
ReliefWeb (2008), Glossary of Humanitarian Terms. http://www.who.int/hac/about/reliefweb-

aug2008.pdf?ua=1
Schoonees, A., Lombard, M. et al. (2013). Ready-to-use therapeutic food for home-based
treatment of severe acute malnutrition in children from six months to five years of age. The
Cochrane Library.
Sphere Project (2011). Humanitarian Charter and Minimum Standards in Emergency

Response. https://www.ifrc.org/PageFiles/95530/The-Sphere-Project-Handbook-20111.pdf
Suri, D.J., Moorthy, D. and Rosenberg, I.H. (2016). The Role of Dairy in Effectiveness and
Cost of Treatment of Children With Moderate Acute Malnutrition A Narrative Review. Food
and Nutrition Bulletin, 37(2), 176-185.
UNHCR, WFP (2011), Guidelines for Selective Feeding: Management of Malnutrition in

Emergencies.
UNICEF (2013) Improving Child Nutrition: The Achievable Imperative for Global Progress
UNICEF (2014). Preventing Moderate Acute Malnutrition (MAM) Through Nutrition Specific
Intervention. CMAM Forum Technical Brief.
UNICEF, WHO and World Bank Group (2012). Levels and Trends in Child Malnutrition, Joint
Malnutrition Estimates 2012 Edition. Washington, DC.
UNICEF, WHO and World Bank Group (2016). Levels and Trends in Child Malnutrition, Joint
Malnutrition Estimates 2016 Edition. Washington, DC.
Wamani, H., Peterson, S. et al. (2007). Boys are more stunted than girls in sub-Saharan
Africa: a meta-analysis of 16 demographic and health surveys. BMC Pediatrics 7, 17.
Webb P., Schlossman, B.L. et al. (2011). Improving the Nutritional Quality of U.S. Food Aid:
Recommendations for Changes to Products and Programs. Boston, MA: Tufts University.
Webb, P., Boyd, E. et al. (2014). Nutrition in emergencies: Do we know what works?. Food
Policy 49, 33-40.
WHO (1999). Management of severe malnutrition: a manual for physicians and other health
workers. Geneva: http://www.scielosp.org/scielo.php?script=sci_arttext&pid=S1020-
49891999000700020
WHO (2000). The management of nutrition in major emergencies. Geneva: WHO.
WHO and UNICEF (2009). WHO Child Growth Standards and the Identification of Severe
Acute Malnutrition in Infants and Children: A Joint Statement.
http://www.who.int/nutrition/publications/severemalnutrition/9789241598163/en/
WHO (2013). Guideline: updates on the management of severe acute malnutrition in infants
and children. World Health Organization.
Published by Oxfam GB under ISBN 978-0-85598-914-9 in March 2017. DOI: 10.21201/2017.9149.
Oxfam GB, Oxfam House, John Smith Drive, Cowley, Oxford, OX4 2JY, UK.
OXFAM
Oxfam is an international confederation of 20 organizations networked together in more than 90 countries, as
part of a global movement for change, to build a future free from the injustice of poverty. Please write to any
of the agencies for further information, or visit www.oxfam.org.

Recovery, Relapse, and Episodes of Default in The Management of Acute Malnutrition in Children in Humanitarian Emergencies: A Systematic Review

Cargado por

Información del documento

Título original

Derechos de autor

Formatos disponibles

Compartir este documento

Compartir o incrustar documentos

Opciones para compartir

¿Le pareció útil este documento?

¿Este contenido es inapropiado?

Copyright:

Formatos disponibles

Recovery, Relapse, and Episodes of Default in The Management of Acute Malnutrition in Children in Humanitarian Emergencies: A Systematic Review

Cargado por

Copyright:

Formatos disponibles

MARCH 2017

RECOVERY, RELAPSE AND EPISODES

About the research team

Copyright Oxfam GB 2017

APPENDIX 1: DATABASE SEARCH STRATEGY STRINGS 15

APPENDIX 2: ASSESSMENT OF INCLUDED STUDIES 17

APPENDIX 3: INCLUDED PUBLICATIONS REPORTED RESULTS 19

Definitions and scope

The research team:

synthesized the evidence in response to two key research questions:

What evidence was eligible for review?

What are the findings?

Study Country Type Quality Findings

Figure 0.2: Reasons for default reported in studies

Study Country Type Quality

offered or not applicable

What is the state of the evidence on acute malnutrition?

The lack of qualitative data on acute malnutrition in humanitarian contexts is striking,

Child-level data, disaggregated by age and gender, is routinely collected in the

Barriers to successful completion of acute malnutrition treatment were multi-factorial and

Defining acute malnutrition in children aged 659 months

Classification Severe acute malnutrition (SAM) Moderate acute malnutrition (MAM)

In the aftermath of a humanitarian emergency, a breakdown in health services, destruction

Management of acute malnutrition

1.2 RATIONALE AND PURPOSE

1.3 RESEARCH QUESTIONS

The review investigated:

0.1 POPULATION, INTERVENTION, CONTROL AND

Relapse rate/repeated relapse episodes: the proportion of children who re-enrolled

Time to recover: the length of time between admission and discharge.

2.2 IMPLEMENTATION OF THE SEARCH STRATEGY

2.3 MODIFICATIONS FROM THE PUBLISHED PROTOCOL

Revisions following peer review

2.4 DOCUMENT SCREENING

2.5 EVIDENCE APPRAISAL

Figure 2.1: CASP checklist questions

2.6 DATA EXTRACTION AND EVIDENCE SYNTHESIS

Figure 3.1: PRISMA diagram of included studies

Academic databases Grey literature Personal contact Other

Inclusion/ 7,591 Articles

Inclusion/ 33 Articles 45 Articles

3.2 CHARACTERISTICS OF INCLUDED STUDIES

Of the 24 studies/programme reports eligible for inclusion, 22 reported on quantitative

3.3 ASSESSMENT OF STUDY QUALITY

3.4 NARRATIVE SYNTHESIS OF RESULTS

The third programme report evaluated four components of an integrated management of

Reasons for default and relapse or return defaults/episodes of default

A programme evaluation report from an outpatient SAM treatment intervention conducted in

4.1 LIMITATIONS OF THE REVIEW

4.2 INFLUENCE OF OPERATIONAL REALITIES ON

4.3 AGE AND SEX DATA

4.4 RECOMMENDATIONS FOR FUTURE RESEARCH

FINAL SEARCH STRATEGY

GREY LITERATURE SEARCH

Exposure accurately measured

Rates or proportions reported

Provides implications for

Can results be applied?

QUALITY ASSESSMENT OF INCLUDED QUANTITATIVE RCTS AND

Groups similar at start of trial

Clinically important outcome

Benefits worth harms/cost

QUALITY ASSESSMENT OF INCLUDED QUALITATIVE STUDIES